Download 17 e. Virtual bond model provides an accurate description of the

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

List of types of proteins wikipedia , lookup

Western blot wikipedia , lookup

QPNC-PAGE wikipedia , lookup

Genetic code wikipedia , lookup

Protein wikipedia , lookup

Biosynthesis wikipedia , lookup

Protein adsorption wikipedia , lookup

Expanded genetic code wikipedia , lookup

Metalloprotein wikipedia , lookup

Intrinsically disordered proteins wikipedia , lookup

Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup

Biochemistry wikipedia , lookup

Protein structure prediction wikipedia , lookup

Transcript
17
e. Virtual bond model provides an accurate description of the conformational preferences of
the backbone
In low resolution approaches, it is important to preserve as much as possible of the unique, diverse
characteristics of different residues, while removing the atomic details. The virtual bond
approximation yields an almost unequivocal description of the backbone conformation, and a
reasonable account of residue specificities. As a matter of fact, the distribution of bond angles and
torsional angles in the virtual bond model are highly correlated with the local secondary structure,
and thus reflect the secondary structure propensities of individual amino acids. See § VII.x.
A total of 2n variables define the backbone geometry for the virtual bond model: the dihedral angle
(ϕi) of each virtual bond, and the angle (ϑi) between successive dihedral angles. The virtual bond
lengths are almost fixed at 3.81 Å, except for cis peptide bonds. The angles ϑi and ϕi follow
interdependent probability distributions specific to each type of residue. Figure II.1.11 displays the
joint distribution of ϕi and ϕi+1. Two regions are highly populated, located near (ϕi, ϕi+1) ≈ (50°,
50°) and ≈ (200°, 200°), characteristic of α-helices, and β-sheet structures, respectively
The distributions NA(ϕi, ϕi+1) for particular amino are illustrated in Figure II.1.12. The strong
preference of Glu, for example, for α-helices, the versatility of Gly to adopt various torsional
states, the aptitude of Val for participating in β-sheets, or the unique rotational preferences of Pro
are distinguishable. The conformational energies extracted from such distributions will be shown
in § VII to be useful for estimating the secondary structure propensities of particular amino acids.
Likewise, the bond angle ϑi is strongly correlated with the torsions ϕi and ϕi+1 of the adjoining
bonds, as illustrated in Figure II.1.13. Bond angles of about 90° and 120° are favored in α-helices
and β-sheets, respectively. The distributions NA(ϑi, ϕi+1) and NA(ϑi, ϕ) for specific residues (not
shown) also exhibit residue-specific characteristics {Bahar, Kaplan, et al. 1997 ID: 92}.
18
Figure II.1.11. Distribution of adjacent virtual bonds' dihedral angles NX(ϕi, ϕi+1). The horizontal axes are the
torsion angles ϕi and ϕi+1, and the surface represents the number of occurrence of each region of size (ϕi ± 15°,
ϕi+1 ± 15°) in 150 PDB structures. The projection of the surface is shown on the lower plane. The most populated
dihedral angle pairs are enclosed by the innermost contours. The sharp peak near (ϕi, ϕi+1) = (50°, 50°)
corresponds to α-helices, and the broader peak near (ϕi, (200°, 200°) is associated with β-sheets. (from {Bahar,
Kaplan, et al. 1997 ID: 92})
19
Figure II.1.12. Number distribution of virtual bond torsions NA(ϕi, ϕi+1) for A = Gly, Glu, Val and Pro.
(taken from {Bahar, Kaplan, et al. 1997 ID: 92})
20
Figure II.1.13. Coupling between virtual bond torsion and bending angles. The surfaces represent the
number distributions from 150 PDB structures, irrespective of residue type. (a) Distribution NX(ϑi,
ϕi+1) for any ith bond angle ϑi and the succeeding bond torsion ϕi+1 (b) Distribution NX(ϑi, ϕi) for
any ith bond angle ϑi and the preceding bond rotation ϕi. Bond angles outside the displayed range 60°
< ϑi < 180° are not displayed. Grids of size 10° were taken in analyzing the bond angles (taken from
{Bahar, Kaplan, et al. 1997 ID: 92})
21
f. Amino acid sidechains prefer angles near their ideal rotational isomeric states
In addition to φ and ψ angles, proteins also have freedom in the side chain rotational
angles. Moving along the side chain away from the backbone defines carbons
identified as Cβ, Cγ, etc., and rotational angles as χ1, χ2, etc.; see Figures II.1.14 and
15.
Hydrocarbon chains [(-CH2-)n] in the gas phase tend to populate the three rotational
isomeric states trans, gauche+ and gauche- (see Figures II.1.5 and 6). Figures II.1.16
and 17 show that the side chains in proteins also tend to populate the same rotational
angles, at least for the χ1 angles. Figure II.1.18 shows that this correspondence
between side chain angles observed in proteins with their ideal values grows stronger as
the structural quality of examined proteins increases. This is usually interpreted to
mean that when proteins are known to very high resolution, side chain angles in
globular proteins should coincide closely with the angles intrinsically favored by those
bond types. The less optimistic interpretation is that protein side chain angles become
ideal because computerized structural refinement methods are based on assuming such
ideality, and protein structure refinements do in part reflect the artifacts of the
refinement process.
But if it is true, observations of φ ψ, and χ angles in proteins indicate an important
principle that holds at least to first approximation: the folding forces acting on proteins
do not perturb the backbone and side chain angles very much relative to the intrinsic
values that amino acids and dipeptides would have in the absence of the folding forces.
Local factors - a side chain interacting with its own backbone, or a methylene group
interacting with a neighboring methylene in a side chain - dictate the bond angle
options that are available to a protein. A folding protein, chooses from among those
options. See Figures II.1.16 and 17. Folding forces rarely distort or strain the angles
the backbone and side chain bonds intrinsically prefer, but simply act to select among
the intrinsically favorable forms.
22
Figure II.1.14. Two examples, Trp and Lys, illustrating the definition of the torsion angles χ1, χ2, etc. of
amino acid side chains in proteins, along with the labels α, β, γ, δ, etc. assigned to successive atoms for
distinguishing their position along the sidechain. Nitrogen atoms are shown in black. Backbone atoms are on
the right end in each case.
23
Figure II.1.15. Definition of sidechain atoms’ identification labels and corresponding torsional
angles, displayed for all types of amino acid. The last column shows the torsional angles
required to be defined in order to fix the positions of all sidechain atoms. See the diagram of
{Ponder 2000 ID: 498}.
24
1000
β- strands
800
600
400
200
0
-180
-120
-60
0
χ1
60
120
180
1600
α -helices
1400
1200
1000
800
600
400
200
0
-180
-120
-60
0
χ1
60
120
180
Figure II.1.16. Distribution of the χ1 sidechain torsion angles for amino acids in α-helical and β-sheet regions
compiled from 150 high resolution protein structures. The ordinate represents the total number of observations for
each interval of size ∆χ1 = 100. Note the relatively low probability of the gauche- state (χ1 = 60 o) for side chains on
α-helices.
25
200
300
Ser
160
Thr
250
120
200
150
80
100
40
50
0
0
-180
-120
-60
0
χ1
60
400
120
180
Ile
-180
150
200
100
100
50
0
0
-120
-60
0
χ
1
60
120
-60
0
χ1
60
200
300
-180
-120
180
120
180
Asn
-180
-120
-60
0
χ1
60
120
Figure II.1.17. Comparison of the χ1 sidechain torsion angles for four different types of amino acids, Ser, Thr, Ile, and Asn,
compiled from 150 high resolution databank structures. The figure illustrates that the side chains choose from among the same
options (about trans, gauche+ and gauche- states) although the relative populations of these states differ depending on the type of
amino acid.
180
26
Figure II.1.18. Deviations of χ1 angles from the standard gauche-, trans and gauche+ rotamers plotted
against the resolution of the experimentally determined protein native structure. Points are for individual
proteins, stars are averages for a given resolution. (adapted from {Thornton 1992 ID: 496})